It might be worth adding to this discussion that Skeptical Inquirer's objective is to call attention to claims of dubious scientific merit so that they can be given further scrutiny. My reading of the article in question is not that it is attacking the methods of econometrics, but rather noting the frequently inconsistent conclusions researchers can reach due to disagreement on what data should be used and when -- they're acknowledging the diamond but making clear its flaws. As such it performs a valuable function; it reminds us that econometrics must be cautiously applied. This is Armchair Economists; Landsburg himself chuckles about how econometrics can be used to show that capital punishment either deters 13 murders per execution or causes 3. I think the SI article makes substantially the same point.
--Brian [EMAIL PROTECTED]